Backyard Aquaponics
A place to discuss aquaponics

After misplacing my API test chart, and often having problems working out my pH level from colours that are very similar I decided that there must be another way. I wondered if I can use my smartphone instead. I seem to be able to use my smartphone for almost everything else, including as a torch.
So I downloaded one of the many free apps that allow the camera on the smartphone to take a “picture” of the colour, and convert it into its colour components. One of the common way of defining colour by its red green and blue components (RGB). This is defined either as 3 separate numbers or a combined 8 digit Hex value, which is written like #45F344. So now I can take a picture of the test tube and get the exact colour. The next step is to compare this with the API test chart.

I created a web site to do this, where you can enter the Hex number or RBG values and the measurement type (pH high or low, Ammonia, Nitrite or Nitrate) and the program will calculate the nearest value to the colour on the API test chart. I have now been using it for a while and like the fact that it takes the guesswork out of working out the values, and it I don’t have to keep looking for my chart.

access it directly onhttp://www.ourresearch.net/api-test-kit-values.html?Next=/my-diary.html. If you keep an on-line diary you can use the “Add to Diary” button to directly add the measurements to your diary.

Damned fine.


I've been taking pics of my test tubes on my phone.
But remember that if you have led or fluro lights the colours in your pic can be changed quite dramatically.
Would you mind telling us what the app you downloaded is called or maybe what to search for in the App Store? Cheers

Once again Sheer brilliance.

I searched the app store for "RBG colour" and there were pages of them. I picked any free one which worked OK so not sure it its the best one. The one I am using is called RGB Colour picker.

I found that its very tricky getting the exact colour and the colour can change a little each time you take it, maybe the test tube is on a slight angle, or you are getting different external light sources. But after a lot of testing and noting that even if the colour didn't seem exactly right or a bit different each time the it small differences didn't make any difference to the result.

It gets the result by calculating the the colour distance between the entered colour and each API colour and picking the closest, which seems quite robust.

Interesting in whether other find this robust as well.

Graphic artists, printers and the sign industry will not use electronic devices for colour matching, no two computer monitors give the same colour results. Mobile phones are the same, the colours don't match from one phone screen to the next even across the same brand of phone.

The only way around the problem is to use the industry standard Pantone Colour Chart.

Colours are referred to by number and are therefore a perfect match each and every time, that's the reason you match your test water sample beside a printed colour sample and not a electronic screen.

The test kit sample colours are all printed to a Patone standard colour and DO NOT vary in colour from one batch of printing to the next and are completely reliable.




http://www.cal-print.com/InkColorChart.htm

Hi
I completely agree that computer screens and electronic devices don't give a perfect colour match. I get this fairly constantly when you present a powerpoint in a seminar the colours you chose on the powerpoint are not at all like those appearing on the big screen.

You can also tell this by using a smartphone to work out the RGB values of the printed chart. They vary slightly each time you take a photo.

What we are looking for here is not the exact colour, but the closest colour on the test chart to the colour in the test tube. This can be calculated by an algorithm. The errors in the electronic devices should be a lot less than the difference between the colours so it shouldn't significantly affect the result.

I've found with the tricky pH values (ie 6.4-7.2) even if the colours that are displayed on the screen on the chart don't look quite like I expected them to, the matching value corresponds to the value I get when I check it on the original. printed chart.

I would be interested in the experiences of other people.

Ria you say you are not looking the exact colour, but the closest colour on the test chart to the colour in the test tube, I don’t know why you would want collect data that is not correct.

You can’t take a photo and expect to get an accurate colour match, there is so much in getting the right lighting, side light, back light, fill, high key, low key all affect the color and density. Different camera positions on the curvature of the test tube will give you different light and a difference in colour.

Even a photo taken outdoors against a photo taken indoors would be different colours, the difference against shaded and non shaded would be different again, you’re talking to an old film lighting guy here.

The pH values of 6.4 – 7.2 are a perfect example of why you need too match to a colour chart and not a photograph that would give you a totally wrong reading more often than not. The colour charts are all printed so they DO NOT vary in colour, the test solution is made to a formula so nothing is left to chance.

Ria your website says that your project is being run as a citizen science project within Macquarie University Australia study and I would have thought that correct data would be paramount in any university research and not use any data that would not be 100% correct.

You said that using your smartphone to work out the RGB values of the printed chart vary slightly each time you take a photo that alone could throw your pH reading out the door and that is the problem. The RGB colour system also has problems with its colours often appearing brighter under different light conditions, again throwing the values out.

Aquarium test kit colour samples are printed from a combination of Pantone colour numbers to give a correct precise and accurate colour.

Comparing your mobile phone screen to a printed chart is not accurate, computers and televisions display images on their screens. RGB colors often appear brighter and more vivid specifically because the light is being projected directly into the eyes of the viewer.

Even Pantone say that their colour chart may not match viewed from a digital screen and to use their printed colour chart.
Pantone® Matching System Color Chart
PMS Colors Used For Printing Labels
________________________________________
This chart is intended as a reference guide only. PANTONE Computer Video simulations displayed may not exactly match PANTONE®-identified color standards. Use current PANTONE Color Publications for most accurate color.

What are PMS colors?
Normally every professional in the premium and gift industry knows what PMS means: Pantone Matching System. Here is the problem at hand: How to explain to a printing company color you want them for printing your business cards or a promotional product ?

Sure, you can send them an image via email with the color, but this image will show up differently on his screen as it did on your own. In fact I work with two screens and when I display the same image on both screen colors will never be perfectly the same.

Every computer monitor is different, every printer is different. Unless your equipment is calibrated with the Pantone hue, the color depicted on your screen will not be accurate and could be many shades off.

This is why the PMS, Pantone matching system was created. It's a standard language for color identification and communication. When you say to the printer : I want you to print a pink 1767C, you can be sure he knows which colour you mean. So it's easy, just use a pantone guide , which contains over 1000 different colours and tell the printer the pantone code of the colour you choose. What if you don't have a pantone guide ? surely, any professional printing company has one, and for sure you can ask them to have a look.

What is RGB Color?
Digital cameras and scanners and create images using combinations of just three colors: Red, Green and Blue (RGB). These are the primary colors of visible light and this how computers and televisions display images on their screens. RGB colors often appear brighter and more vivid specifically because the light is being projected directly into the eyes of the viewer.

This is an "additive" process in which the three colors are combined in different amounts to produce various colors. It is called "additive" because you must add varying amounts of two or more colors to achieve hues and values other than the three basic red, green and blue colors.

Computer monitors and televisions vary the amount of each color from 0 to a maximum of 255. Equal maximum amounts of all three colors (often expressed as R255, G255, B255) creates white. The absence of all three colors (R0, G0, B0) creates black. Equal amounts of all three colors somewhere between 0 and 255 will create varying shades of gray.

That's interesting. As I said. Taking a pic with your phone will be affected by the surrounding light.
But for the purpose of our testing we are not comparing the colour from our screen, the computing system that is looking at the raw data is picking the numbers then comparing those numbers to the numbers of the test chart, and giving up a result of the nearest match.
I'm trusting that against my failing eyesight when I can't tell the difference between the colours on the chart let alone the colour between the chart and the test tube.
When I take a pic of my tests on my phone I put the test tube on the colour I think it matches best then click and look on the screen I often find that I'm a colour match out which with PH means 7.4 to 8 could be the difference between minor inconvenience to a HSM.
I trust that little electric eye more than mine.
But remember we are not trying to get the colour of the pic on the screen to match the chart. We are getting the numbers of the pic to match the numbers of the chart and I think that is what Ria was trying to get across.

The interesting result of this discussion is that it should not be a big internet search to get a ph colour listing by numbers / presuming the chemical tests are all the same. That means it would be possible for any printing company to print a chart with more variants on it.

That's interesting. As I said. Taking a pic with your phone will be affected by the surrounding light.

Yes and it's the result of that photo that you are taking your information from, take a second photo from a different angle and you will get a different result from the first.


But for the purpose of our testing we are not comparing the colour from our screen, the computing system that is looking at the raw data is picking the numbers then comparing those numbers to the numbers of the test chart, and giving up a result of the nearest match.


It doesn't matter what is seen on the screen it's the image you take that gives you your data and that can vary with different light or by using different camera angles so how can you get correct data to compare to anything?



When I take a pic of my tests on my phone I put the test tube on the colour I think it matches best then click and look on the screen I often find that I'm a colour match out which with PH means 7.4 to 8 could be the difference between minor inconvenience to a HSM.


That is the reason why you can't match it digitally and there are too many variables to get a correct reading.



But remember we are not trying to get the colour of the pic on the screen to match the chart. We are getting the numbers of the pic to match the numbers of the chart and I think that is what Ria was trying to get across.


Eric don't worry about what's on the screen it's the photo that will vary and throw the whole match out so it doesn't matter what numbers you get, just take a second photo from a different angle and the colour reflected back into the lens will have a different colour value. So what happens then, split the difference between the two numbers or just pull numbers out of a hat?



When I take a pic of my tests on my phone I put the test tube on the colour I think it matches best then click and look on the screen I often find that I'm a colour match out which with PH means 7.4 to 8 could be the difference between minor inconvenience to a HSM.


This is the difference between matching by eye and what the camera picks up and that is what I'm saying.



The interesting result of this discussion is that it should not be a big internet search to get a ph colour listing by numbers / presuming the chemical tests are all the same. That means it would be possible for any printing company to print a chart with more variants on it.


All test kits are not the same and possibly aren't the same chemicals I use both API and Sera test kits, the both use the same 5mls. volume of water and one uses 3 drops of solution and the other use 4 so they can’t be the same chemical or each test kit would have a different reading to the other.

With the API kit testing both pH and pH high using two different chemicals, the Sera chemical covers both ranges using only one solution, another chemical entirely so your chemical tests aren’t all the same and your printing company can print as many variants it wants but they wont match different brands of test kits and chemicals.

Even the media each test company uses to print their colour samples on will always be printed on the same stock so the test colour sample doesn't vary.

It’s not just a matter of choosing a colour it must be matched correctly to the corresponding test solution to give the correct and accurate reading anything other than that precise match is INCORRECT




I forgot to say in my previous post what I was quoting came from this site below and was not just my personal advise.
http://www.ginifab.com/feeds/pms/

A really interesting discussion, mostly beacause you are all correct in one way or another.

Joblow - kudos to you. You are precise and accurate in your assertions. However, you run the risk of being overly so in what is a process with an inherrant error. Colour matching will always have a +/- that will vary with the eye that beholds. As such, nothing can be 100% correct there will always be a confidence associated with any measure.

Ria, as a piece of citizen science - nice work!

The point I think, and I think this is what Ria was seeking feedback on, is the reliabilty of the results in real terms. That is, if the errors associated with the phone taking a picture of the test results and the subsequent analysis do not alter the assessment of pH, which is a range in itself, then there is no harm done.

This can only be done by sampling, which Ria is seeking help with from us. How robust is the phone method with different phones as compared with the eye doing the matching? Great question.

Ria, good luck mate.

Carbonfarmer why would you add more variables than you need too, using camera just adds more problems to the mix and more chance of error. Why try too re-invent the wheel?


This site may explain it a little clearer.

http://www.cambridgeincolour.com/tutori ... tering.htm


As quoted:
The whitest regions are those which contribute most towards the exposure calculation, whereas black areas are ignored.

Meaning that if your test vial does not completely fill the entire picture frame the background colour is influencing your result and any stray light hitting the vial just changes the result even further.

Again another variation that is the reason looking down on your test vial from above on a white background with no direct light influencing the colour is the CORRECT way to match your colour sample.

Maybe I'm missing something, If you guys can explain to me why you would want anything other than a CORRECT reading I'll be very happy to take it on-board.

Am I the only one who feels like Ria is being picked on a bit?

Actually this may assist me, if I can ever work out how to use the dratted mobile phone. I am colour blind and usually have to get someone else to interpret the results.

Nothing more embarrassing than when I was younger buying a light grey shirt and finding out it was pinkish.

I like the idea of the app. Make it work and try to sell it to API

One suggestion to curb the error/camera range. Add a calibration setting for the photo/image. So take a picture of a test tube, and then manually eyeball the test tube and input it into the phone. So the app will know how much it differs with the picture taken from a normal camera.

But I think it'd be pretty close to accurate as is. Would be quite handy if you could make it bring up the level on the phone as well, so dark purple comes up with 5ppm nitrite.

I'd use it if I could just take a photo of all 4 tests at once, and it'd display the results automatically. Then record the date and time it was taken.

It could be done like this, have your colour samples printed 100mm x 100mm with a 25mm white diam. circle in the centre and a test tray 25mm diam. and have your solution sit over directly over the white circle. You would still need to stick with the 5mls. of water as the kit requires.

Printer’s can measure the ink density and print a perfect colour match from your test card colour samples but it would need too be a correct match.

The solution needs to be viewed against the white with no other colour overlaping, any overlap would add another colour.

You would need to take your photo from directly above and completely frame your picture within the 100mm x 100mm square that way you would not have any other colour influencing the calculation.

The low test tray would pick up very little side light and because you’re not shooting from side on there would be no curvature to pick up stray light and throw the colour out. Take the photo in subdued light.